Re: Splitting into several docs; more extensive use/emphasis of profiles

["" <kaeo@merike.com>]

Hi George.

OK...this is right up my alley and I will take a look at the profiles
section next week and see what modifications I can suggest.  Please forward
any other comments you personally get regarding the profiles to the list or 
me so I can try and incorporate.  Do you still want me to "own" the profile
section for editing?  What is your timeframe for trying to have the next
rev ready?

- merike


----- Original Message -----
From: George M. Jones
Sent: 6/11/2003 8:22:55 PM
To: opsec@ops.ietf.org
Subject: Splitting into several docs; more extensive use/emphasis of profiles

> [Someone not on the list] wrote:
> 
> > At 10:43 PM 6/11/2003, you wrote:
> >
> >> [Someone not on the list] wrote:
> >>
> >>>
> >>> Your document has some significant flaws. Example: You appear to 
> >>> require the presence of MPLS (2.13.1). This may not have been your 
> >>> intent, but it's what's there. I really don't want MPLS running on 
> >>> my laptop. Your document says elsewhere these are requirements for 
> >>> hosts, routers, and other devices.
> >>>
> >>> This could be remedied by saying "For devices implementing MPLS, 
> >>> ..." or some such.
> >>>
> >>> Please reread your document thinking about what types of devices 
> >>> REALLY need to meet each of the items you suggest.
> >>>
> >>> Your document is really more oriented toward routers, and at that 
> >>> CORE routers.
> >>
> >>
> >>
> >> It admitedly has baggage that shows its heratage (UUNET's 
> >> requirements for core and egdge devices); you've hit on one of them.
> >> The profiles mechanism is is intended to be a vehicle for grouping 
> >> things appropriatly.
> >>
> >>> I think it'd be a good idea to do a fair bit of rework to make 
> >>> clearer what's really important, and why, rather than trying to 
> >>> dictate the feature set of all devices on the Internet.
> >>>
> >>> Keep in mind a host, or a $60 NAT box running at someone's house or 
> >>> small office is just not going to be an appropriate target of your 
> >>> document.
> >>
> >>
> >>
> >> I think a pass needs to be made aimed at making each of the 
> >> requirements more standalone and then emphasizing the use of profiles
> >> to determine different sets of needs (core, edge, toaster...)
> >
> >
> > That'd work.
> >
> > Might also make sense to split this into a family of documents, which 
> > likely would improve readability. 
> 
> 
> I've considered that. It might also solve the BCP vs. Non-BCP problem. 
> Also, might help with editing. Size and information density
> are starting to make it hard to edit/modify already. I'll put it on the 
> list of possible changes for -01.
> 
> 
> Thanks,
> ---George
> 
> 
> 
> 
> 
>